2MASS Point Spread Function
Validation

This material concerns the validation of PSFs based on examination of data from production processing. Data are given for two distinct sets of PSFs, the first for the initial northern set (generated in the winter of 1997-1998) used for the processing of northern data before 980605, the second for the set (generated in late 1998) for the summer 1998 interval, when the telescope was recollimated, starting 980605 and running until (but not including) 980919. All plots are PostScript, and unfortunately come out upside down; click on 'swap landscape' in the orientation menu to see them properly. I apologize for the inconvenience this entails. For each scan, six plots are given for each of the J, H, and Ks bands.

1. Reduced chisq of profile fit vs profile fit magnitude
2. Profile fit uncertainty vs profile fit magnitude
3. Histogram of profile fit sqrt(red. chisq).
4. (Profile fit magnitude - standard aperture mag) vs scan Y, PSF segment boundaries and PSF IDs marked.
5. Reduced chisq of profile fit vs scan Y, PSF segment boundaries and PSF IDs marked.
6. Profile fit magnitude vs standard aperture magnitude

NB., for #3 the histogram of the square root of chisquare is used because it is more nearly normally distributed than the histogram of chisquare, for large degrees of freedom.

For the first interval from spring 1997 until 980605, plots from three scans, 074, 109, and 131 from the RTB (and first Sampler data release) night, 971116n, are given:

971116n
Scan 074: J H K
Scan 109: J H K
Scan 131: J H K

For the 980605n - 980919n interval, scans 068 and 076 are given for 980610n (good seeing). For this period, the PSF ID numbers marked on plots #4 and #5 (vs scan y coord) are related to the seeing (shape) by dividing by 10,000.

980610n
Scan 068: J H K
Scan 076: J H K

and scans 035 and 064 for 980714n (poor seeing):

980714n
Scan 035: J H K
Scan 064: J H K.

A number of phenomena are apparent in the above plots. Since they were made from the raw PROPHOT output .ps files, the (profile fit mag - std. aperture mag) histories (the nominal difference is -1.5 mag) show step offsets with PSF ID change that should be corrected later in MAPCOR. The data from 980714n j064 are especially spectacular for this effect. However changes within a given PSF ID segment are not correctable. Narrow features in these plots, in which profile fit magnitudes are a whole magnitude or more fainter that normal, seem to be mostly due to the presence of extremely bright stars in the images.